Re: [PATCH 6/8] add the functions for cnd_t

[Jens Gustedt <jens.gustedt@inria.fr>]

[-- Attachment #1: Type: text/plain, Size: 2904 bytes --]

Am Samstag, den 30.08.2014, 20:35 -0400 schrieb Rich Felker:
> On Sat, Aug 30, 2014 at 08:47:22PM +0200, Jens Gustedt wrote:
> > diff --git a/src/thread/cnd_destroy.c b/src/thread/cnd_destroy.c
> > new file mode 100644
> > index 0000000..7c24d1a
> > --- /dev/null
> > +++ b/src/thread/cnd_destroy.c
> > @@ -0,0 +1,5 @@
> > +#include <threads.h>
> > +
> > +void (cnd_destroy)(cnd_t *cnd) {
> > +	/* For private cv this is a no-op */
> > +}
> 
> Is there a reason for the parens around the name? I'm assuming it was
> related to a macro definition at some point, but we don't have one now
> and it seems unlikely that it would be desirable to define away the
> destroy call to a NOP, in case we ever want to switch to an
> implementation where NOP is not suitable.

you are assuming correctly, I'll eliminate these

> > diff --git a/src/thread/cnd_init.c b/src/thread/cnd_init.c
> > new file mode 100644
> > index 0000000..c8aaee5
> > --- /dev/null
> > +++ b/src/thread/cnd_init.c
> > @@ -0,0 +1,9 @@
> > +#include <threads.h>
> > +
> > +int cnd_init(cnd_t * c)
> > +{
> > +	*c = (cnd_t) {
> > +		0
> > +	};
> > +	return thrd_success;
> > +}
> 
> Just style, but generally we use {0} or { 0 } all on one line as a
> zero-initializer. I would refrain from using multi-line unless (1)
> we're using designated initializers, not just a universal zero
> initializer, and (2) multi-line is actually needed for clarity or to
> avoid exceeding 80 columns.

yes, this time its not me but astyle that produced that. I was just
tired of cleaning up behind it.

> > diff --git a/src/thread/cnd_signal.c b/src/thread/cnd_signal.c
> > new file mode 100644
> > index 0000000..143883c
> > --- /dev/null
> > +++ b/src/thread/cnd_signal.c
> > @@ -0,0 +1,9 @@
> > +#include <threads.h>
> > +
> > +int __private_cond_signal(cnd_t *, int);
> > +
> > +int cnd_signal(cnd_t * cnd) {
> > +	/* This internal function never fails. */
> > +	(void)__private_cond_signal(cnd, 1);
> 
> Cast-to-void isn't used in musl source except when it's necessary for
> types (e.g. in ?:).

No problem with me, I just thought that there are still compilers out
there that bark on that.

> > +	return thrd_success;
> > +}
> 
> This would be a candidate for "return thrd_success?thrd_success:ret",
> I think.

Hm, so this would make this a tail call. On the other hand as it is
now this states more clearly that we expect the return value of
__private_cond_signal to be irrelevant.

Ok, I'll go with the tail call and put a comment about that
expectation.

Jens

-- 
:: INRIA Nancy Grand Est ::: AlGorille ::: ICube/ICPS :::
:: ::::::::::::::: office Strasbourg : +33 368854536   ::
:: :::::::::::::::::::::: gsm France : +33 651400183   ::
:: ::::::::::::::: gsm international : +49 15737185122 ::
:: http://icube-icps.unistra.fr/index.php/Jens_Gustedt ::



[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 198 bytes --]